Attractive Train Interiors: Minimizing Annoying Sound and Vibration Ulf Carlsson Ulf Orrenius Stockholm 2013 KTH Railway Group and Bombardier Transportation Sweden ISSN 1651‐7660 TRITA‐AVE 2013:28 KTH Railway Group, publication 13‐01 Authors: Ulf Carlsson KTH Aeronautical and Vehicle Engineering SE‐100 44 Stockholm [email protected] and Ulf Orrenius Bombardier Transportation Sweden SE‐721 73 Västerås [email protected] SUMMARY Rail vehicle passengers demand comfortable journeys. A passenger who wishes to work or read during her or his journey needs to be able to focus without being distracted by disturbing sounds or vibrations that makes writing difficult. In addition to the direct disturbance effect, such sounds and vibrations significantly affect passengers’ perception of product quality and are therefore important factors to attract and keep passengers from other less energy effective modes of transportation. In this perspective the acoustic and vibrational interior comfort of rail vehicles is an important factor when seeking to promote travel with relatively low energy and environmental impact. A study of annoying sounds and vibrations generated by train interiors is reported. A number of different types of annoying sounds are discussed with respect to the effects they have on the passengers and a notation for distinguishing annoying sounds of different character is defined. Annoying sounds in vehicles are categorized with respect to the underlying generation mechanisms and measures for mitigation are discussed in general terms as well as the state‐of‐art regarding metrics for analysis of disturbing sounds. Furthermore, a literature survey of annoying sounds and vibrations in cars is presented together with procedures and methodologies to reduce the occurrence of such sounds. It is suggested that pro‐active methodologies to minimize annoying sound and vibration in cars could be transferred and adapted to be used in rail vehicle design and manufacturing, for example component testing in shaker rigs. An investigation of disturbing sounds and interior vibrations on Swedish intercity trains is also reported. It is found that a large majority of the annoying sounds onboard a Swedish intercity train is of tapping and rattling type, originating from components like ceiling panels, light covers, cabinet doors, interior sliding doors and foldable tables. A number of case studies are presented based on observations on operating vehicles. From the survey it is found that for some vehicles the number of annoying sounds and vibration issues related to interiors is substantial. Also for vehicles with less than 10 year operation. This observation underlines the need for systematic abatement procedures and proactive activities from the manufacturers to ensure comfortable train journeys. Finally, best practice design solutions to reduce interior vibrations and annoying sounds from train interiors are presented. The solutions discussed include: • Monitoring and reporting programs in operating vehicles. • Systematic application of vibration testing in the component and system quality assurance programs. • Effective source isolation systems for important vibrating systems like compressors and propulsion systems. • Squeake and rattle free mounting techniques for interior panels, doors and lighting system. • Low vibration design and mounting strategies for passenger chairs and interior tables. Contents Page 1 INTRODUCTION 1 1.1 Disturbing sound and vibrations in rail vehicle interiors 1 1.2 Purpose of investigation 1 2 ANNOYING SOUNDS IN VEHICLES 2 2.1 Sounds caused by impact forces 3 2.1.1 Generating mechanisms 3 2.1.2 Measures to reduce impact generated sound 4 2.2 Sounds caused by friction forces 4 2.2.1 Generating mechanisms 4 2.2.2 Measures to reduce friction generated sound 5 2.3 Objective metrics for annoying sounds 6 2.3.1 Loudness 6 2.3.2 Tonality 6 2.3.3 Sharpness 6 2.3.4 Fluctuation strength and roughness 6 2.3.5 Combined features –Acoustic annoyance indices 7 2.3.6 Objective BSR metrics in automobile industry 7 3 HOW IS ANNOYING SOUND PROBLEMS TREATED IN CAR INDUSTRY? 8 3.1 Present situation 8 3.2 Component vibration testing 8 3.2.1 Component annoying sound detection 9 3.2.2 Component annoying sound diagnosis 9 3.3 End‐of‐production‐line‐inspection 9 3.4 Trends for future 10 4 ANNOYING VIBRATION – VIBRATION COMFORT 11 4.1 How is passenger vibration discomfort assessed? 12 4.2 Vibration effects on passenger activities 13 i 5 ANNOYING INTERIOR SOUND AND VIBRATION IN SWEDISH INTERCITY AND REGIONAL TRAINS 20 5.1 Measurements and analysis 20 5.2 Measurement data analysis 20 5.2.1 Knocking sound 16 5.2.2 Rattling sound 19 5.2.3 Rubbing sound 21 5.3 Interior vibration 25 5.3.1 Table vibration 25 6 MEASURES TO REDUCE ANNOYING INTERIOR FURNISHING SOUND AND VIBRATION 28 6.1 Lessons from automotive industry 28 6.1.1 Material combinations at component interfaces 29 6.1.2 Component vibration testing 31 6.2 Design solutions in the railway industry: Best practice and outlook 33 7 CONCLUSIONS 36 REFERENCES 38 ii 1 INTRODUCTION 1.1 Disturbing sound and vibrations in rail‐vehicle interiors According to the Merriam Webster dictionary 0 the term “noise” is defined as ”any sound that is undesired or interferes with one's hearing of something”. In this perspective sounds of relatively low energetic level can be very annoying. Such sounds and, in addition, disturbing vibrations of rail vehicle interiors are in focus in the present report. Rail vehicle passengers demand comfortable journeys. A passenger who wishes to work or read or just relax during her or his journey needs to be able to focus without being distracted by disturbing sounds or vibrations that makes writing, reading or relaxing difficult. In addition to the direct disturbance effect, such sounds and vibrations significantly affect the perception of product quality Error! Reference source not found. and are an important factor to attract and keep passengers from other less energy effective modes of transportation, such as air‐ and car travel. In this perspective the acoustic and vibrational interior comfort of rail vehicles is an important factor to promote travel with relatively low energy and environmental impact Error! Reference source not found.. Regarding interior vibrations it has been shown that the present international standards for rating vibration comfort do not effectively assess the ability for passengers to write on board rail vehicles. In a field study performed on three Swedish Inter‐City trains Error! Reference source not found. it was found that about two thirds of the 330 randomly selected passengers reported to experience moderately or much difficulty in performing a short written test. In all three trains the measured vibration levels were found to be acceptable according to the applicable international comfort vibration standards 0. 1.2 Purpose of investigation The purpose of the present work is three‐folded. The first aim is to illustrate and analyze typical sound and vibration phenomena with a high degree of annoyance observed in Swedish InterCity trains. The second aim is to survey methods and procedures applied in the automotive industry to eliminate Buzz Squeak and Rattle (BSR) sounds and also other sounds like knocking sounds. Finally, the third aim is to discuss annoying sound and vibration on rail vehicles in a design context, in particular addressing design measures and maintenance procedures to avoid the occurrence of annoying sound and vibration on railway vehicles. 1 2 ANNOYING SOUNDS IN VEHICLES What kind of sounds do we consider as annoying? This question is very complex and is in itself a research topic in psychoacoustics. The complexity is due to the fact that the perception of sound is different from person to person. Even a single individual perceives sounds differently depending on situation and mood. Familiar sounds are less annoying than unfamiliar. Sounds we connect with unpleasant experiences are usually more annoying. A sound in one particular context is more annoying than in a different context. Expectation is yet another factor influencing our perception. If we have bought an expensive car we expect it to be free of sounds indicating low quality like for instance a rattling instrument panel. Finally, to make it even more complicated, it has been found that our reactions depends on how the acoustic disturbance is combined with other types of disturbances for instance vibration, light and heat. What kind of annoying sounds and vibration do we find in vehicles? As passengers we become annoyed when the general noise level is so high that we have difficulties in communicating with other passengers. We become annoyed when suddenly appearing sounds make us lose focus on a demanding task, on reading or writing or a discussion with friends. Hence in vehicles we find both stationary and non‐stationary annoying sounds. Ventilation noise and rolling noise from the wheel‐rail interaction are examples of stationary sounds that can be annoying if they are sufficiently loud. Squealing brakes, or a door bouncing against its doorpost, are examples on non‐stationary annoying sounds. Most people find non‐stationary sounds more annoying than stationary since we have an ability to adapt to stationary sounds. In this investigation we focus on non‐stationary sounds caused by interior furnishing. When we discuss annoying sounds we need to share a common notation to facilitate discussion. A large variety of terms, often sound‐mimicking, have been developed to characterize annoying sounds. The term buzzing sound is used for low‐frequency sounds often radiated from resonantly vibrating surfaces, like the 50 Hz plus harmonics sound generated by electric lights. With knocking or tapping sounds we mean a sound consisting of distinct, often regularly repeated, series of soft impacts, like when we knock on a door or a table. Rattle is a more complex sound consisting of a series of more or less random impacts followed by reverberant sound with wide frequency content, like that generated when shaking steal beads in a tin can. Squeal is a high frequency, typically 600 Hz – 2000 Hz, tonal sound with long duration, like a squealing railway wheel during braking or curving. Squeak is a short duration high frequency tonal sound, like when the rubber sole of a shoe rubs a polished floor. A hissing sound is a long duration high frequency broad‐band sound, like when gas flows out from a small leak. Scratch is a short duration high frequency sound, like two sandpapers in sliding contact. Grunt, hum and moan are all low frequent sounds, typically from say 100 Hz to say 500 Hz, like a door slowly turning on its hinges. And so on … A conclusion drawn by several researchers, see for instance reference [6], is that the probability of annoying sounds to appear in a built‐up structure increases in proportion to the structural complexity and the number of components. A second conclusion is that most of the 2 annoying sound problems are directly related to the product assembly process. Hence to reduce the probability of annoying sounds to be generated in built‐up structures, like vehicles, the designers have to work with the component design and, possibly even more important, with the methods and processes used to assemble the components. In the following sections we discuss annoying sounds in vehicles based on the character of the generating mechanisms. This basis of characterization is motivated by the fact that successful noise abatement procedures are dependent on well understood generating mechanisms. 2.1 Sounds caused by impact forces Common for many of the annoying sounds we find in vehicles is that they are caused by interacting components in relative motion. Tapping, knocking and rattling sounds are caused by components vibrating with a relative displacement directed towards each other. The moving part in a sliding door interacting with its carbody‐fixed door‐frame is one example. If the sliding door vibrates with an amplitude larger than the clearance to the door frame it will repeatedly bounce between the fixed stops and the impacts may generate a disturbing tapping sound. 2.1.1 Generating mechanisms As described above, two adjacent vibrating components will impact each other repeatedly if their relative motion has a component directed normal to the components’ surfaces, and if the clearance between the components is small enough. If the vibration amplitude is sufficiently large the two components will repeatedly impact each other and excite vibration that will radiate a tapping sound. Typically one of the components vibrates in resonance with amplitudes larger than the clearance (gap) between the components. The impacting components’ total motion will, in the linear regime, be a superposition of the original vibration and the vibration response caused by the impacts. The impact force duration and magnitude influence the vibration response of the impacting components. An impact with long duration will excite primarily low‐frequency vibrations whereas short duration impacts excite vibrations also at high frequencies. In the linear regime the responding vibration magnitude is proportional to the impact force amplitude. The dynamic properties of the impacting components also have large influence on the response to the impacts. Lightly damped vibration modes sounds different than strongly damped modes. Also, the position of the resonance frequencies along the frequency axis has large influence on how we perceive the sound generated. Tapping and rattling sounds are both generated by impact forces. The question is then ‐ what are the differences between tapping and rattling sound? One difference is that tapping sounds are clearly distinguishable from each other whereas the impacts forming rattle merge into each other. Another difference is that tapping sounds appear as relatively damped in contrast to rattles that often appear reverberant. Finally the energy contents of rattling sounds are generally higher at high frequencies than that of tapping sounds. There is of course a transition zone in between tapping and rattling sounds where classification is difficult. 3 2.1.2 Measures to reduce impact generated sound When the generating mechanisms are understood various measures to avoid impact generated sounds can be suggested in general terms. • Avoid or reduce the clearances between components. Note that the impacts may also be eliminated with a larger clearance. • Reduce the vibration amplitude to a value lower than the clearance. • Reduce the mass of the impacting component. • Change to a softer material at the impact point. This will shift the frequency contents of the impact sound to lower, less annoying, frequencies. A reduced clearance means that the impact velocity, and hence the impact force, is reduced. Alternatively, the impacts can be eliminated if the clearance is made larger than the largest possible vibration amplitude. Vibration amplitudes can be reduced in several ways. If the vibrations are resonant, either the system losses can be increased or the system eigen‐frequency can be shifted away from the exciting frequency. The idea of changing the contact surface is to shape the spectrum of the radiated sound to one that is less disturbing. Basically this means that the major part of the acoustic energy is shifted to low frequencies where the human auditory system is less sensitive. 2.2 Sounds caused by friction forces Squeal, squeak, squelch and moan, in contrast to tapping and rattling sounds, are all caused by two components in sliding contact, i e with parallel relative motion components. When the components slide over each other a time‐varying friction force will excite vibration that will generate sound of different types. Time‐varying friction forces can be caused by various phenomena. The most common is known as stick‐slip but a time‐varying normal force or contact area also cause time‐ varying friction forces. Some researchers have classified stick‐slip generated sounds as either squealing sounds or rubbing sounds. Squealing sounds are of tonal character and rubbing sounds are of broadband character. 2.2.1 Generating mechanisms Sound from friction is a complex area and research is still performed to better understand and describe its generation mechanisms, see for instance references [7 – 11]. Stick‐slip is a phenomenon where the sliding component repeatedly sticks and slips on the contact surface. Suppose two components, one deformable and one for simplicity rigid, are connected over a surface. If one of the components, for instance the deformable, starts to move relative to the other a friction force sticking the contacting surfaces together develops in the contact surface. The friction force balances the spring force caused by the deformation of the deformable component. As the displacement of the deformable component increase the spring force also increase. This continues until a point where the static friction force has reached its maximum value. When the spring force increases above the maximum static friction force, the components starts to slide and the friction force drops 4 sharply to the lower dynamic friction force. Since the spring force is now larger than the counteracting friction force the contact surfaces will slide with an accelerating speed until a point where the friction force again balances the spring force. When the spring force is smaller than the friction force the sliding speed will decrease to a point where the surfaces stick again. Then the spring force starts to increase again and the stick‐slip process repeats. The period of the stick‐slip repetition is determined by the several factors, • the relative component speed, • the friction coefficient’s speed‐dependence, • the normal force in the contact surface and • the stiffness and mass properties of the deformable component. In cases where the stick‐slip frequency is sufficiently close to one of the component’s eigen‐ frequencies the stick‐slip deformation amplitude will grow large and possibly cause radiation of tonal sound. In reality both components are deformable meaning that structural modes of any of the interacting bodies may be excited by the forces generated. However, the basic description above still holds. Also, the process described above is a repeated stick‐slip motion superimposed over a relative motion of the two components. The relative motion serves as a energy reservoir supplying the sound generating mode(‐s) with energy. When the supplied energy is balanced by the dissipated energy a steady‐state vibration is reached. In a vehicle the overall relative motion is typically caused by either one of the two components vibrating relative to its equilibrium position. In the case described above the stick‐slip phenomenon gave rise to a series of periodic forces which excited a resonant vibration and a tonal squealing or squeaking sound. In practice the complexity of the interacting surfaces may be such that the stick‐slip forces will have a more complex character and the vibration and sound generated will be more or less random in character, i e a scratching sound. As a model for this generation process one can imagine a sandpaper in contact with a rough sandy surface. When the surfaces with their randomly distributed grits slide over each other the asperities will stick and slip randomly. The vibration and sound generated will be similar to squeak but instead of tonal character it will have high‐frequency random character. 2.2.2 Measures to reduce friction generated sound From the list of factors influencing the stick‐slip motion we can find some measures to reduce or even prevent its appearance. • Avoid contact between components if not necessary. • Prevent relative motion between the components. • Change friction coefficient characteristics by surface treatments or change of materials, see Section 6.1.1 below for further information. • Change dynamic properties of components to avoid locking to a structural vibration mode. • Increase the normal force. 5
Description: